Multiple items
val string : value:'T -> string

Full name: Microsoft.FSharp.Core.Operators.string

--------------------
type string = System.String

Full name: Microsoft.FSharp.Core.string
Multiple items
val float : value:'T -> float (requires member op_Explicit)

Full name: Microsoft.FSharp.Core.Operators.float

--------------------
type float = System.Double

Full name: Microsoft.FSharp.Core.float

--------------------
type float<'Measure> = float

Full name: Microsoft.FSharp.Core.float<_>
Multiple items
val int : value:'T -> int (requires member op_Explicit)

Full name: Microsoft.FSharp.Core.Operators.int

--------------------
type int = int32

Full name: Microsoft.FSharp.Core.int

--------------------
type int<'Measure> = int

Full name: Microsoft.FSharp.Core.int<_>
type 'T list = List<'T>

Full name: Microsoft.FSharp.Collections.list<_>
module Option

from Microsoft.FSharp.Core
union case Option.Some: Value: 'T -> Option<'T>
union case Option.None: Option<'T>
Multiple items
type Async
static member AsBeginEnd : computation:('Arg -> Async<'T>) -> ('Arg * AsyncCallback * obj -> IAsyncResult) * (IAsyncResult -> 'T) * (IAsyncResult -> unit)
static member AwaitEvent : event:IEvent<'Del,'T> * ?cancelAction:(unit -> unit) -> Async<'T> (requires delegate and 'Del :> Delegate)
static member AwaitIAsyncResult : iar:IAsyncResult * ?millisecondsTimeout:int -> Async<bool>
static member AwaitTask : task:Task -> Async<unit>
static member AwaitTask : task:Task<'T> -> Async<'T>
static member AwaitWaitHandle : waitHandle:WaitHandle * ?millisecondsTimeout:int -> Async<bool>
static member CancelDefaultToken : unit -> unit
static member Catch : computation:Async<'T> -> Async<Choice<'T,exn>>
static member FromBeginEnd : beginAction:(AsyncCallback * obj -> IAsyncResult) * endAction:(IAsyncResult -> 'T) * ?cancelAction:(unit -> unit) -> Async<'T>
static member FromBeginEnd : arg:'Arg1 * beginAction:('Arg1 * AsyncCallback * obj -> IAsyncResult) * endAction:(IAsyncResult -> 'T) * ?cancelAction:(unit -> unit) -> Async<'T>
static member FromBeginEnd : arg1:'Arg1 * arg2:'Arg2 * beginAction:('Arg1 * 'Arg2 * AsyncCallback * obj -> IAsyncResult) * endAction:(IAsyncResult -> 'T) * ?cancelAction:(unit -> unit) -> Async<'T>
static member FromBeginEnd : arg1:'Arg1 * arg2:'Arg2 * arg3:'Arg3 * beginAction:('Arg1 * 'Arg2 * 'Arg3 * AsyncCallback * obj -> IAsyncResult) * endAction:(IAsyncResult -> 'T) * ?cancelAction:(unit -> unit) -> Async<'T>
static member FromContinuations : callback:(('T -> unit) * (exn -> unit) * (OperationCanceledException -> unit) -> unit) -> Async<'T>
static member Ignore : computation:Async<'T> -> Async<unit>
static member OnCancel : interruption:(unit -> unit) -> Async<IDisposable>
static member Parallel : computations:seq<Async<'T>> -> Async<'T []>
static member RunSynchronously : computation:Async<'T> * ?timeout:int * ?cancellationToken:CancellationToken -> 'T
static member Sleep : millisecondsDueTime:int -> Async<unit>
static member Start : computation:Async<unit> * ?cancellationToken:CancellationToken -> unit
static member StartAsTask : computation:Async<'T> * ?taskCreationOptions:TaskCreationOptions * ?cancellationToken:CancellationToken -> Task<'T>
static member StartChild : computation:Async<'T> * ?millisecondsTimeout:int -> Async<Async<'T>>
static member StartChildAsTask : computation:Async<'T> * ?taskCreationOptions:TaskCreationOptions -> Async<Task<'T>>
static member StartImmediate : computation:Async<unit> * ?cancellationToken:CancellationToken -> unit
static member StartWithContinuations : computation:Async<'T> * continuation:('T -> unit) * exceptionContinuation:(exn -> unit) * cancellationContinuation:(OperationCanceledException -> unit) * ?cancellationToken:CancellationToken -> unit
static member SwitchToContext : syncContext:SynchronizationContext -> Async<unit>
static member SwitchToNewThread : unit -> Async<unit>
static member SwitchToThreadPool : unit -> Async<unit>
static member TryCancelled : computation:Async<'T> * compensation:(OperationCanceledException -> unit) -> Async<'T>
static member CancellationToken : Async<CancellationToken>
static member DefaultCancellationToken : CancellationToken

Full name: Microsoft.FSharp.Control.Async

--------------------
type Async<'T>

Full name: Microsoft.FSharp.Control.Async<_>
static member Async.Parallel : computations:seq<Async<'T>> -> Async<'T []>
val async : AsyncBuilder

Full name: Microsoft.FSharp.Core.ExtraTopLevelOperators.async
Multiple items
type MailboxProcessor<'Msg> =
  interface IDisposable
  new : body:(MailboxProcessor<'Msg> -> Async<unit>) * ?cancellationToken:CancellationToken -> MailboxProcessor<'Msg>
  member Post : message:'Msg -> unit
  member PostAndAsyncReply : buildMessage:(AsyncReplyChannel<'Reply> -> 'Msg) * ?timeout:int -> Async<'Reply>
  member PostAndReply : buildMessage:(AsyncReplyChannel<'Reply> -> 'Msg) * ?timeout:int -> 'Reply
  member PostAndTryAsyncReply : buildMessage:(AsyncReplyChannel<'Reply> -> 'Msg) * ?timeout:int -> Async<'Reply option>
  member Receive : ?timeout:int -> Async<'Msg>
  member Scan : scanner:('Msg -> Async<'T> option) * ?timeout:int -> Async<'T>
  member Start : unit -> unit
  member TryPostAndReply : buildMessage:(AsyncReplyChannel<'Reply> -> 'Msg) * ?timeout:int -> 'Reply option
  ...

Full name: Microsoft.FSharp.Control.MailboxProcessor<_>

--------------------
new : body:(MailboxProcessor<'Msg> -> Async<unit>) * ?cancellationToken:System.Threading.CancellationToken -> MailboxProcessor<'Msg>
static member MailboxProcessor.Start : body:(MailboxProcessor<'Msg> -> Async<unit>) * ?cancellationToken:System.Threading.CancellationToken -> MailboxProcessor<'Msg>
val printfn : format:Printf.TextWriterFormat<'T> -> 'T

Full name: Microsoft.FSharp.Core.ExtraTopLevelOperators.printfn
namespace System
type Int32 =
  struct
    member CompareTo : value:obj -> int + 1 overload
    member Equals : obj:obj -> bool + 1 overload
    member GetHashCode : unit -> int
    member GetTypeCode : unit -> TypeCode
    member ToString : unit -> string + 3 overloads
    static val MaxValue : int
    static val MinValue : int
    static member Parse : s:string -> int + 3 overloads
    static member TryParse : s:string * result:int -> bool + 1 overload
  end

Full name: System.Int32
System.Int32.TryParse(s: string, result: byref<int>) : bool
System.Int32.TryParse(s: string, style: System.Globalization.NumberStyles, provider: System.IFormatProvider, result: byref<int>) : bool
val sin : value:'T -> 'T (requires member Sin)

Full name: Microsoft.FSharp.Core.Operators.sin
val cos : value:'T -> 'T (requires member Cos)

Full name: Microsoft.FSharp.Core.Operators.cos
Multiple items
module List

from Microsoft.FSharp.Collections

--------------------
type List<'T> =
  | ( [] )
  | ( :: ) of Head: 'T * Tail: 'T list
  interface IEnumerable
  interface IEnumerable<'T>
  member GetSlice : startIndex:int option * endIndex:int option -> 'T list
  member Head : 'T
  member IsEmpty : bool
  member Item : index:int -> 'T with get
  member Length : int
  member Tail : 'T list
  static member Cons : head:'T * tail:'T list -> 'T list
  static member Empty : 'T list

Full name: Microsoft.FSharp.Collections.List<_>
val fold : folder:('State -> 'T -> 'State) -> state:'State -> list:'T list -> 'State

Full name: Microsoft.FSharp.Collections.List.fold

Introduction to F#

The Turtle Tutorial

About Me

Luke Merrett

Technology Manager at JUST EAT

Co-Organiser of F# |> Bristol and Big Data Bristol

https://lukemerrett.com

@LukeAMerrett

What is F#?

F# is a strongly typed functional programming language built atop the .NET CLI

Originally developed by Don Syme from Microsoft Research, it has the conciseness of Python, strictness of Scala & ecosystem of .NET.

Benefits

Taken from Scott Wlaschin's summary here

The "5Cs" of F#...

Conciseness

F# gets rid of a lot of the clutter / noise, removing curly brackets, semi colons etc:

1: 
2: 
3: 
4: 
5: 
6: 
7: 
8: 
9: 
// Functions are very light weight
let add x y = x + y

// This is a Record Type declaration; an immutable DTO
type Person = {First:string; Last:string}

// Here we create an instance of the Person type
// It's type is inferred by the property names
let bob = {First="Bob"; Last="Baggings"}

A lot of this is due to a powerful type inference system the compiler uses than can work out static types without having to explicitly declare them.

Convenience

The lightweight nature of F# makes common programming tasks simpler. This is especially true for rapidly building domain models.

 1: 
 2: 
 3: 
 4: 
 5: 
 6: 
 7: 
 8: 
 9: 
10: 
11: 
12: 
13: 
// Low overhead type definitions
type Coord = {Lat:float; Long:float}

type TimePeriod = Hour | Day | Week | Year
type Temperature = C of int | F of int
type Appointment = OneTime of DateTime | Recurring of DateTime list

// "Railway orientated" concise, easy to read chains of logic
config
|> TableManager.createTable client
|> TableManager.waitForTableCreation client
|> DataManager.streamRowsFromSourceFile
|> DataManager.loadRowsIntoDynamo client

As functions are first class citizens, the emphasis is on functional composition of reusable code over inheritance.

Correctness

Values are immutable by default, nulls are replaced with Option types, and it supports explicit units of measure, preventing a large class of errors.

 1: 
 2: 
 3: 
 4: 
 5: 
 6: 
 7: 
 8: 
 9: 
10: 
11: 
let bob = {First="Bob"; Last="Baggings"}
bob.First <- "Jim" // Assignment error

// Option types explicitly capture that a value may not be present
// These can be thought of like Nullable<T> types in .Net
let getResponseBody: string Option = 
    match response.StatusCode with
    | 200 -> Some response.Body
    | 404 -> None

let distance = 10<m> + 10<ft> // errors, can't add meters to feet

As F# is strongly typed, you get all the compile time benefits of .Net without the overhead.

Concurrency

F# has a set of built in libraries for asynchronous logic; not only with the async {} wrapper but also a built-in actor model.

1: 
2: 
3: 
4: 
5: 
6: 
7: 
8: 
Async.Parallel [ for i in 0..40 -> async {
    return fib(i)
}]

MailboxProcessor.Start(fun inbox-> async {
    let! msg = inbox.Receive()
    printfn "message is: %s" msg
}) 

This is enhanced by the focus on immutable data structures which allow sharing of state & avoid locking of resources.

Completeness

Whilst F# focused on being functional, it is a multi-paradigm language that supports OO if required.

As it is built ontop of the .Net CLI, it has full access to the existing .Net ecosystem, NuGet packages and C# assemblies.

1: 
2: 
3: 
4: 
5: 
6: 
7: 
8: 
9: 
// Values can be made mutable if necessary
let mutable counter = 0

// .Net libaries can be used seamlessly
open Newtonsoft.Json
let person = JsonConvert.DeserializeObject<Person>(content)

// Out parameters are actually handled in a cleaner way
let (successfully, parsedValue) = System.Int32.TryParse("123");

Supported in both .Net 4.5+ and .Net Core. Has tooling for Visual Studio 2012 onwards & Visual Studio Code

The Exercise

We'll be working through an F# script that models a Logo Turtle, allow us to give it basic commands such as Move 10, Turn Left 90, Pen Down and Set Colour Green.

The script is incomplete; as we walk through the workshop we will be adding to the code until it is a functional Logo program.

At the end we'll pass a series of commands to our Turtle and check it responds as expected.

Exercise Setup

Clone the F# Turtle Tutorial

Walk through the README.md file to get F# running locally

Then open ./Exercise/TurtleRunner.fs

The FSI

The F# tools you've installed include a REPL called "FSI". This will let us execute each part of the tutorial without having to build each time, whilst allowing us to execute individual lines or sections of code on the fly.

To use this in VS Code, select the lines you wish to execute and press Alt+Enter. This will send the code to FSI and run it, giving you back the declared types and output.

Don't select the module TurtleRunner line when using FSI, it won't be able to interpret it.

Basic Discriminted Unions

1: 
type Animal = Cat | Dog | Fish | GuineaPig

Otherwise known as option types; in the most basic sense these can be thought of as enums.

Fill in the exercise script so we have:

  • Pen states of "Up" and "Down"
  • Colours of "Black", "Red", "Blue" and "Green"
  • Turn Directions of "Left" and "Right"

Record Types

1: 
type Person = {firstName:string; lastName:string; age:int}

Record types can be thought of as sealed classes with a set of read only properties.

They act like value types, as in when equality is checked the value of all properties is used, not the object reference.

Fill in the Turtle record so it has these properties:

  • "xpos" - float type
  • "ypos" - float type
  • "angle" - float type
  • "penState" - PenState type
  • "colour" - Colour type

Significant Whitespace

F#, just like Python, uses significant whitespace for indentation.

1: 
2: 
3: 
4: 
5: 
6: 
7: 
8: 
9: 
// Totally cool
type Person = {
    firstName:string
    lastName:string}

// Not cool, won't compile
type Person = {
    firstName:string
lastName:string}

Line Endings & Semi Colons

You'll also see some of the examples use ; and some use new lines. The two are synonymous with each other

1: 
2: 
3: 
4: 
5: 
6: 
7: 
8: 
// If on one line, you need to use ";"
type Person = {firstName:string; lastName:string; age:int}

// But on multiple lines, the new line acts as the terminator
type Person = {
    firstName:string
    lastName:string
    age:int}

The Algebraic Type System

Think of the type system as set logic rather than defining objects

Sum types are "Discriminated Unions", Product types are "Tuples"

 1: 
 2: 
 3: 
 4: 
 5: 
 6: 
 7: 
 8: 
 9: 
10: 
11: 
12: 
13: 
14: 
15: 
16: 
17: 
// Sum type - denoted by "|"
// This type contains a sum of all the possible values
type Speed = Slow | Fast    // 2 possible values
type Direction = Up | Down  // 2 possible values

// Product type - denoted by "*"
// Contains:
//  - A set of all possible Speeds 
//  - TIMES
//  - A set of all possible Directions
type Movement = Moving of Speed * Direction 

// The Product of our 2 sum types is 4 possible values
Moving(Slow, Up)
Moving(Slow, Down)
Moving(Fast, Up)
Moving(Fast, Down)

Discriminated Unions with Values

Here's the true power of discriminated unions; they allow us to declare values that must be provided to initialise them.

1: 
2: 
3: 
4: 
type Shape =
    | Circle of int 
    | Rectangle of int * int
    | Point of int * int

To initialise a Rectangle we'd then write Rectangle(40,20).

Fill in the Command type so it has these choices:

  • Move which takes in a float
  • Turn which takes a TurnDirection and a float
  • SetPen which takes in a PenState
  • SetColour which takes in a Colour

Functions

1: 
let add x y = x + y

Functions in F# are extremely light weight, this is as the compiler can infer types for most cases.

Here note as well we aren't saying return explicitly; instead F# always returns the last value in the function.

1: 
let add (x:int) (y:int): int = x + y

Above is an example with explicit declarations of the types expected. The signature would be int -> int -> int with the last type being the return type.

Pattern Matching

We've got a set of commands which are of the discriminated union type Command; so how do we decompose the values based on the type of command?

1: 
2: 
3: 
4: 
5: 
match command with
    | Move distance -> printfn "move %f" distance
    | Turn(direction, degrees) -> printfn "turn %A %f" direction degrees
    | SetPen state ->  printfn "set pen to %A" state
    | SetColour colour -> printfn "set colour to %A" colour

Pattern matching allows you to check type and bind the values inside that type in one step. Think switch statements on crack.

Apart from being succinct, one of the core benefits of pattern matching is that the compiler warns you if any of the possible states have been missed.

Copy this code into the processCommand function for now.

Immutability and the "With" Syntax

We have the processCommand function with the signature Turtle -> Command -> Turtle; so we know we need to return a modified turtle based on the command we have received (e.g: the turtle has moved 10 places).

However F# types are immutable by default, so we need a way to create a new Turtle record type to pass back.

1: 
{turtle with xpos=20, ypos=10}

The with syntax in F# lets us do this without copying each value manually to a new record type, here it's saying:

"Create a new record type instance using all fields in turtle but change xpos and ypos"

Implementing Process Command

Below is the logic you will need to move the turtle based on its current angle.

1: 
2: 
3: 
4: 
let angleInRads = turtle.angle * (Math.PI/180.0)
{turtle with 
    xpos = turtle.xpos + distance * sin angleInRads
    ypos = turtle.ypos + distance * cos angleInRads}

Copy this in after | Move distance -> in the processCommand function.

Once you add this the other match statements will go red; this is because Move distance is now returning a Turtle type, but the other statements are still returning a unit type.

Now finish implementing the rest of the pattern match expressions for Turn, SetPen and SetColour

Using the Process Command

Now we've implemented the solution, we can uncomment all code under:

1: 
// --- Uncomment this section to run a full test ---

Note how block comments in F# use (* *) instead of /* */

Lists

The below logic creates a List<Command> to test our domain and function.

 1: 
 2: 
 3: 
 4: 
 5: 
 6: 
 7: 
 8: 
 9: 
10: 
11: 
let commands = [
    Move 20.0
    Turn(Left, 90.0)
    Move 20.0
    Turn(Right, 90.0)
    SetColour Red 
    Move 20.0
    Turn(Right, 90.0)
    SetPen Up 
    Move 40.0
]

Creating Record Types

This creates our initial turtle record:

1: 
let turtle = {xpos=0.0; ypos=0.0; angle=90.0; penState=Down; colour=Black}

Each value of the record type must be supplied, there are no nulls.

Note how we've not specified the name of the record type being instantiated, the F# compiler infers this from the property names

List.fold

This is like a reduce operation; it takes in an initial value (turtle) and a list to iterate over (commands).

1: 
2: 
3: 
List.fold (fun agg command -> processCommand agg command) 
          turtle 
          commands

For each item in the list, it passes in the current aggregator (agg) and the current list item (command), applies a function to them which returns a new aggregator (so here, a new Turtle type).

In our context; this boils down to taking a turtle and a command, processing that command, then returning a turtle with the updated state so we can apply the next command

Pipe Forward Operator

The pipe forward operator |> takes the preceding value and puts it into the last argument of the following function.

1: 
2: 
3: 
let movedTurtle = 
    commands 
    |> List.fold (fun agg command -> processCommand agg command) turtle

This is really syntax sugar however leads to code that states it's intentions clearly, such as

1: 
2: 
3: 
4: 
File.ReadAllText "appsettings.json" 
|> JsonConvert.DeserializeObject<AppSettings> 
|> ConnectToDatabase
|> GetUserList 

Testing the Logic

In VS Code, select the whole file except the module TurtleRunner line and press Alt+Enter to send to F# Interactive

If the exercise is completed, this should run without error returning a Turtle in the state:

1: 
{xpos = 40.0; ypos = -20.0; angle = 180.0; penState = Up; colour = Red}

Where Next?

If you'd like to continue learning F#: